faintly chiming on a still Sabbath morning, and it may be doubted whether any offering has, of recent years, been left at their 'parlour' on Sheepstor, though there was a time when no one was supposed to visit this hollow in the rocks without depositing at least a pin as a present to ensure their goodwill. The farmer who returns from Ashburton Market after his usual hour no longer dares to lay the blame on 'they piskies,' as heretofore, for no one will believe him. The pixies are departing from Dartmoor. Like beings of larger growth, they do not like expressions of incredulity as to the feats they have accomplished; and so they have retired among the granite boulders for aye, and are no more seen.

CHAPTER III.

ANTIQUITIES.

List of Antiquities-Hut-Circles-Pounds-Trackways-Tracklines-Barrows and Cairns-Time of their Construction-Cromlechs-Kistvaens-Sacred Circles-Menhirs-Stone Avenues-Ancient Bridges-Crosses.

BARREN, solitary, desolate as it is now, Dartmoor is rich in traces of former inhabitants. Scarcely a hillside but has its hut-circle; scarcely a river that shows not the ruins of buildings connected with ancient streaming' operations. Fragments of trackways frequently lost in the bogs, tracklines marking out divisions of aboriginal villages, are ever and anon encountered on steep slope or wind-swept plain, while barrows and cairns, crowning its hills, mark the resting-place of mighty ones long since passed away.

Then there are those mysterious circles of upright stones which have been generally, though, with little doubt, erroneously, looked upon as places of worship; the still more mysterious menhir or rock pillar, the cromlech, and the avenues or parallelitha, whose use has probably given rise to more controversy than all. Can it be marvelled at that these strange relics of the past should invest Dartmoor

with a charm all their own?-a charm almost as powerful as, though quite distinct from, that feeling of weirdness which cannot fail to be inspired in the soul of every thoughtful lover of nature in her wilder aspects who shall tread its wastes.

First of the hut-circle and pound. As we have endeavoured to show, at a very remote period a considerable population worked the surface of the Moor for tin, and the remains of their rude circular dwellings-usually a mere fringe of granite blocks-and their walled enclosures, or 'pounds,' point to this occupation. It is difficult-in fact, impossible-to discriminate between the huts of the ancient Britons, who sought the white metal in the time of Diodorus, and those of a much later date, for miners continued to 'stream' to an appreciable extent as late as the reign of Elizabeth. But it has been considered, and that by no incompetent authority, that ' only those which occur in the immediate neighbourhood of sacred circles, of ancient graves, and, perhaps, of stone avenues, may safely be referred to the British period.*

It is not, of course, necessary to assign all the ancient dwellings on the Moor to those engaged in mining operations. Doubtless, many thousands of sheep and cattle were depastured there then as now, whose owners erected for themselves and their families these rude shelters from the blast. The wild hunter, too, had his home upon the forest; and when the sun had sunk to rest beyond the distant line of heath, would return to the rude shieling, and placing his celt† in its accustomed place, stretch his limbs to the blaze of the peat fire which burned in thè middle of the floor, ere, enveloping himself in his wolfskin trophy, he sought a healthful repose. To these savage fastnesses, too, as we have seen, came the British *R. J. King, p. 11, note.

From the Latin celtis, a chisel. Vide vol. i. of Thomas Hearne's edition of Leland, Appendix, sub tit. A Discourse concerning some Antiquities found in Yorkshire' (Wright's 'Celt, Roman and Saxon,' p. 73).

warrior, flying from his Saxon foe, as the Bretwalda ever advanced westward. Here was he secure from his enemies, and hence might he still harry them by sudden descents into the lowlands, to disappear again as quickly among the rocky piles and inaccessible morasses of his mountain home.

These huts appear to have been sometimes conical in shape, sometimes of beehive construction. They were built of undressed stone without cement, the interstices stopped with turf, and with an inner lining of thin slabs. From the spring' of the roof it seems probable that either reeds or wattles were commonly used, the aboriginal architect seldom being sufficiently skilled to complete a dome of the same materials as those employed in the construction of the walls. Smoke escaped through an aperture in the centre, and, with the exception of what entered at the door, the dwelling was without light. The entrances almost invariably faced south. These primitive habitations are generally found on the hillside, and within easy reach of a stream, and are more numerous on the borders than in the interior. They vary much in size, having diameters from twelve to thirty-six feet, and can thus have scarcely been used as sheepfolds, as is the opinion of some writers. Occasionally much larger examples occur, which may have been devoted to the protection of cattle, but it is just as probable that they were occupied by the chieftain of the community.

The pound was the proper fold for the wild man's flocks and herds, though it probably served a double purpose, and was converted into a refuge for the clan in times of war or when danger menaced. Here and there we find a village built within the enclosure, of which the finest specimen is Grimspound, which we shall visit in our exploration of the east quarter. A pound of another shape, and the only example, was some years since to be found at Manaton, and here, instead of the roughly piled

but massive walls, a large elliptical enclosure constructed of immense blocks of granite, set closely together, and standing erect, is said to have presented itself. This circumvallation, as compared with others, was on low ground, and, in the opinion of Mr. Rowe, designed for cattle only.

Trackways were roads, and, perhaps, also boundaries. They were distinctly marked on the surface of the Moor by being paved, though irregularly, with stones, and must have proved the very roughest of thoroughfares. Remains of these roads occur in several places, but the most lengthy portion now extant is to be found near Longaford Tor, and is supposed to have led across the forest in an almost direct line from Hameldon to Great Mis Tor, connecting, it is presumed, Grimspound with the settlement above Merivale Bridge. It may be seen in high preservation coming down the northern slope of Chittaford Down towards the banks of the East Dart. Here it can be traced for a considerable distance, and is visible running due west through Hollocombe, and up the opposite hill to Little White Tor. Down the common towards the Dart it bends towards the north-east, but in the level, near Postbridge, it takes a direction southward.'* Occasionally peat-cutters have encountered it covered over by the encroaching vegetation.

Oftentimes low banks of earth and stone are observed among the traces of ancient settlements. These are tracklines. They mark not only the enclosures pertaining to one or more hut-circles, but, according to Sir Richard Colt Hoare, were also for communication and defence. The design is usually rectangular, but they are not always found in this form; on Cosdon Beacon they are curvili

* Rowe, p. 46. Some consider this trackway to be the remains of the old Roman Fosseway. This, however, seems to have run from Exeter to Totnes by the present Ridgway, thence through Plympton and Tamerton into Cornwall. Vide R. J. King, p. 132, and vide Appendix C.